| View previous topic :: View next topic |
| Author |
Message |
Goverp
Advocate
Joined: 07 Mar 2007
Posts: 2202
|
 Posted: Thu Jun 25, 2020 11:57 am Post subject: Error 255 in kernel 5.7 compiling arch/x86/events/amd/ibs.o |
 |
|
Ever since kernel 5.7.0, it won't compile. (This is gentoo-sources). The error is at the end of the following listing, I've included the lines before in case the context is important. I copied my 5.6.x .config and ran "make oldconfig" beforehand.
| Code: | make clean
CLEAN arch/x86/entry/vdso
CLEAN arch/x86/realmode/rm
CLEAN arch/x86/tools
hp /usr/src/linux-5.7.6-gentoo # make -j1
HOSTCC scripts/basic/fixdep
HOSTCC arch/x86/tools/relocs_32.o
HOSTCC arch/x86/tools/relocs_64.o
HOSTCC arch/x86/tools/relocs_common.o
HOSTLD arch/x86/tools/relocs
...
CC arch/x86/events/amd/core.o
arch/x86/events/amd/core.o: warning: objtool: amd_pmu_handle_irq()+0x28: stack state mismatch: cfa1=7+8 cfa2=7+0
CC arch/x86/events/amd/uncore.o
CC arch/x86/events/amd/power.o
CC arch/x86/events/amd/ibs.o
arch/x86/events/amd/ibs.o: warning: objtool: perf_ibs_init()+0x25: can't find jump dest instruction at .text+0x37
make[3]: *** [scripts/Makefile.build:267: arch/x86/events/amd/ibs.o] Error 255
make[3]: *** Deleting file 'arch/x86/events/amd/ibs.o'
make[2]: *** [scripts/Makefile.build:488: arch/x86/events/amd] Error 2
make[1]: *** [scripts/Makefile.build:488: arch/x86/events] Error 2
make: *** [Makefile:1732: arch/x86] Error 2
|
AFAIK the warning messages are irrelevant; my 5.6 kernel produces tons of them during compilation.
I suspect the problem may be down to some kernel reorganization of the AMD IBS code that I can't now locate. What I don't understand is why I'm (almost) the only user in the world hitting this (almost - someone seems to have reported it against genkernel.
_________________
Greybeard |
|
| Back to top |
|
 |
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6214
Location: Dallas area
|
| Posted: Thu Jun 25, 2020 12:02 pm Post subject: |
|
|
Do you have the kernel set for a particular CPU? If so which one?
And which gcc are you using?
_________________
UM780, 6.12 zen kernel, gcc 13, openrc, wayland |
|
| Back to top |
|
|
Goverp
Advocate
Joined: 07 Mar 2007
Posts: 2202
|
| Posted: Thu Jun 25, 2020 12:16 pm Post subject: |
|
|
Anon-E-moose, good call! It was my KCFLAGS="-march=native" that caused the problem.
_________________
Greybeard |
|
| Back to top |
|
|
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6214
Location: Dallas area
|
| Posted: Thu Jun 25, 2020 12:39 pm Post subject: |
|
|
I've seen mixed results with "native", better to set the cpu chip directly, and if there isn't one available, then use the next best option (prev version cpu)
_________________
UM780, 6.12 zen kernel, gcc 13, openrc, wayland |
|
| Back to top |
|
|
Ionen
Developer
Joined: 06 Dec 2018
Posts: 2892
|
| Posted: Thu Jun 25, 2020 2:09 pm Post subject: |
|
|
| Considering the kernel goes out of its way to disable a lot of -m options (such as -mno-avx, likely for a reason).. but not every single that exists (especially newer things), and -march=native enables those, I imagine it could have all sort of strange implications. |
|
| Back to top |
|
|
Jimmy Jazz
Guru
Joined: 04 Oct 2004
Posts: 333
Location: Strasbourg
|
| Posted: Fri Jun 26, 2020 12:30 am Post subject: |
|
|
| Ionen wrote: | | Considering the kernel goes out of its way to disable a lot of -m options (such as -mno-avx, likely for a reason).. but not every single that exists (especially newer things), and -march=native enables those, I imagine it could have all sort of strange implications. |
it is the missing -mno-tbm flag
_________________
« La seule condition au triomphe du mal, c'est l'inaction des gens de bien » E.Burke
| Code: |
+----+----+----+
| |::::| |
| |::::| |
+----+----+----+ |
motto: WeLCRO
WritE Less Code, Repeat Often |
|
| Back to top |
|
|
|
Kernel & Hardware
